Re: Is replacing transactions with CTE a good idea? - Mailing list pgsql-general

From Dave Cramer
Subject Re: Is replacing transactions with CTE a good idea?
Date
Msg-id CADK3HHKcZ09TLNNUvcm_++NwDQeHR=J=Qx=oZwayP5e0ZDweTw@mail.gmail.com
Whole thread Raw
In response to Re: Is replacing transactions with CTE a good idea?  (Bruce Momjian <bruce@momjian.us>)
Responses Re: Is replacing transactions with CTE a good idea?  (Bruce Momjian <bruce@momjian.us>)
List pgsql-general


On Thu, 1 Apr 2021 at 15:39, Bruce Momjian <bruce@momjian.us> wrote:
On Thu, Apr  1, 2021 at 11:24:48AM -0400, Dave Cramer wrote:
> CTE's don't change the isolation level. I'm not sure what you are getting at
> here ?

I think what he/she means here is that all queries in a CTE use a single
snapshot, meaning you don't see changes by commits that happen between
queries that are part of the same CTE.  If you were running the queries
separately in read committed mode, you would see those changes, but you
would not see them in repeatable read or serializable transaction mode.

OK, that makes sense, but I think it is wrong minded to think that this absolves one of taking isolation into account.

When you make the first read you will still have to deal with all of the isolation issues 


Dave Cramer
www.postgres.rocks

pgsql-general by date:

Previous
From: Ron
Date:
Subject: Re: Slick way to update multiple tables.
Next
From: Bruce Momjian
Date:
Subject: Re: Is replacing transactions with CTE a good idea?